3D network with channel excitation and knowledge distillation for action recognition

Abstract

Modern action recognition techniques frequently employ two networks: the spatial stream, which accepts input from RGB frames, and the temporal stream, which accepts input from optical flow. Recent researches use 3D convolutional neural networks that employ spatiotemporal filters on both streams. Although mixing flow with RGB enhances performance, correct optical flow computation is expensive and adds delay to action recognition. In this study, we present a method for training a 3D CNN using RGB frames that replicates the motion stream and, as a result, does not require flow calculation during testing. To begin, in contrast to the SE block, we suggest a channel excitation module (CE module). Experiments have shown that the CE module can improve the feature extraction capabilities of a 3D network and that the effect is superior to the SE block. Second, for action recognition training, we adopt a linear mix of loss based on knowledge distillation and standard cross-entropy loss to effectively leverage appearance and motion information. The Intensified Motion RGB Stream is the stream trained with this combined loss (IMRS). IMRS surpasses RGB or Flow as a single stream; for example, HMDB51 achieves 73.5% accuracy, while RGB and Flow streams score 65.6% and 69.1% accuracy, respectively. Extensive experiments confirm the effectiveness of our proposed method. The comparison with other models proves that our model has good competitiveness in behavior recognition.

DILS: depth incremental learning strategy

“…During the training process, the student model attempts to replicate the predictions of the teacher model in order to transfer its knowledge. This method can reduce the size of the model, thereby improving its efficiency (Lan et al, 2022 ; Hu et al, 2023 ). Model pruning (Han et al, 2015 ) is a technique of compressing the model's size by removing unimportant neurons or connections.…”